Case studies using OPERA data

[kvenkman]

Here is a list of events that can be explored as part of the tutorials to be presented at SciPy. The idea is to have 6 events in all to present on - 3 in each category. The events have to have occurred post 03/23 for DSWx and post 01/22 for DIST to be made currently available through OPERA.

Surface Water Extent:

1. Lake Mead water levels over 2023 - bounced back after significant dip during 2021-2023
2. Texas flooding (May 2024) - can show before/after images to estimate flood extent area
3. Annual variations in Vaigai Reservoir (India) - can show time series to calculate reservoir area, potentially volume

Disturbance:

1. Wildfire event in Greece (currently written up, using DIST-ALERT)
2. Deforestation in Brazil (state of Para, can be seen in DIST-ANN product)
3. The Great Green Wall (Senegal, (16.50, -16.09)) - variations visible in the DIST-ALERT product

[dhavide]

@kvenkman this looks great.

I'd prefer having these six examples spelled out more explicitly & quantitatively (i.e., specific spatial coordinates rather than "Greece" or "Texas" & a time window of appropriate length rather than "2023"). Naively, I think of an "event" as a box containing a point in space-time. That is:

• a rectangle in some spatial coordinates e.g., [($\phi_1$, $\theta_1$), ($\phi_2$, $\theta_2$)], the lower-left & upper-right corners of the box in lat-lon or UTM coordinates in a suitable zone).
• an interval in time $[T_1, T_2]$ for two time-stamps $T_1$ & $T_2$.

Can you add something like that to each of the six items above? These don't need to be super precise, just enough to define a suitable superset of the relevant space-time region.

• It may be preferable to specify the spatial AOI using a single geographic point and, say, a radius (which would give a space-time circular cylinder rather than a box). Regardless, this doesn't have to be super precise; the goal is to give users some numbers from which they can set up an initial search in the EarthData database for later refinement.
• I realise we can't count on fine-grained temporal resolution due to period of satellite orbits (as well as limited spatial resolution). These limitations are something we'll have to explain to users at the tutorial. Regardless, let's encourage users to map "events of interest" to some kind of "space-time box" as a mental picture to start with.